DRAFT VERSION DECEMBER 7, 2017Typeset using LATEX twocolumn style in AASTeX61

COL-OSSOS: COLORS OF THE INTERSTELLAR PLANETESIMAL 1I/‘OUMUAMUA

MICHELE T. BANNISTER,1 MEGAN E. SCHWAMB,2 WESLEY C. FRASER,1 MICHAEL MARSSET,1 ALAN FITZSIMMONS,1

SUSAN D. BENECCHI,3 PEDRO LACERDA,1 ROSEMARY E. PIKE,4 J. J. KAVELAARS,5, 6 ADAM B. SMITH,2 SUNNY O. STEWART,2

SHIANG-YU WANG,7 AND MATTHEW J. LEHNER7, 8, 9

1Astrophysics Research Centre, School of Mathematics and Physics, Queen’s University Belfast, Belfast BT7 1NN, United Kingdom2Gemini Observatory, Northern Operations Center, 670 North A’ohoku Place, Hilo, HI 96720, USA3Planetary Science Institute, 1700 East Fort Lowell, Suite 106, Tucson, AZ 85719, USA4Institute for Astronomy and Astrophysics, Academia Sinica; 11F AS/NTU, National Taiwan University, 1 Roosevelt Rd., Sec. 4, Taipei 10617, Taiwan5Herzberg Astronomy and Astrophysics Research Centre, National Research Council of Canada, 5071 West Saanich Rd, Victoria, British Columbia V9E 2E7,

Canada6Department of Physics and Astronomy, University of Victoria, Elliott Building, 3800 Finnerty Rd, Victoria, BC V8P 5C2, Canada7Institute of Astronomy and Astrophysics, Academia Sinica; 11F of AS/NTU Astronomy-Mathematics Building, Nr. 1 Roosevelt Rd., Sec. 4, Taipei 10617, Taiwan8Department of Physics and Astronomy, University of Pennsylvania, 209 S. 33rd St., Philadelphia, PA 19104, USA9Harvard-Smithsonian Center for Astrophysics, 60 Garden St., Cambridge, MA 02138, USA

(Received 2017 November 16; Revised 2017 December 4; Accepted 2017 December 6)

Submitted to ApJL

ABSTRACT

The recent discovery by Pan-STARRS1 of 1I/2017 U1 (‘Oumuamua), on an unbound and hyperbolic orbit, offers a rare oppor-tunity to explore the planetary formation processes of other stars, and the effect of the interstellar environment on a planetesimalsurface. 1I/‘Oumuamua’s close encounter with the inner Solar System in 2017 October was a unique chance to make observationsmatching those used to characterize the small-body populations of our own Solar System. We present near-simultaneous g′, r′, andJ photometry and colors of 1I/‘Oumuamua from the 8.1-m Frederick C. Gillett Gemini North Telescope, and gri photometry fromthe 4.2 m William Herschel Telescope. Our g′r′J observations are directly comparable to those from the high-precision Coloursof the Outer Solar System Origins Survey (Col-OSSOS), which offer unique diagnostic information for distinguishing betweenouter Solar System surfaces. The J-band data also provide the highest signal-to-noise measurements made of 1I/‘Oumuamua inthe near-infrared. Substantial, correlated near-infrared and optical variability is present, with the same trend in both near-infraredand optical. Our observations are consistent with 1I/‘Oumuamua rotating with a double-peaked period of 8.10±0.42 hours andbeing a highly elongated body with an axial ratio of at least 5.3:1, implying that it has significant internal cohesion. The color ofthe first interstellar planetesimal is at the neutral end of the range of Solar System g − r and r − J solar-reflectance colors: it is likethat of some dynamically excited objects in the Kuiper belt and the less-red Jupiter Trojans.

Keywords: minor planets, asteroids: individual (1I/2017 U1)

Corresponding author: Michele T. Bannistermichele.t.bannister@gmail.com

arX

iv:1

711.

0621

4v2

[as

tro-

ph.E

P] 6

Dec

201

7

2 BANNISTER ET AL.

1. INTRODUCTION

The first detection of an interstellar minor planet, 1I/2017U1 (‘Oumuamua), came on 2017 October 19, at mV = 19.6,by Pan-STARRS1 (Chambers et al. 2016)) (Bacci et al. 2017;Meech et al. 2017). In 19 years of digital-camera all-skysurveying, it is the first definitively interstellar, decameter-scale object to be found. The earlier lack of detections hasimplied a low density of interstellar planetesimals (Francis2005; Cook et al. 2016; Engelhardt et al. 2017). Several Earthmasses of ejected bodies are expected per star during plane-tary formation and migration (e.g. Levison et al. 2010; Bar-clay et al. 2017). Given the number of Galactic orbits sincethe major ejection of planetesimals from the Solar System,1I/‘Oumuamua is statistically unlikely to originate from theSolar System.

1I/‘Oumuamua’s orbit1 has a securely extrasolar origin.1I/‘Oumuamua came inbound to the Sun on a hyperbolic andhighly inclined trajectory that radiated from the solar apex,with an orbital eccentricity of e = 1.1994± 0.0002 and in-clination of 122.7◦, avoiding planetary encounters. Suchorbits are not bound to our Solar System. The planetes-imal was travelling at a startlingly high velocity of v∞ =26.02 ± 0.40 km/s. This velocity is typical for the meanGalactic velocity of stars in the solar neighborhood (Mama-jek 2017).

The physical properties of 1I/‘Oumuamua are not yet wellconstrained. Its approach geometry and fast passage left onlya brief window when it was observable, after its 0.16 auminimum approach to Earth on 2017 October 14. As theNEOWISE scans missed its outbound trajectory (J. Masiero,pers. comm.), no albedo is yet known. 1I/‘Oumuamua hasan absolute magnitude of HV = 22.08 ± 0.452, implying asize of . 200 m, assuming it has an albedo in the rangeseen for either carbonaceous asteroids or Centaur albedos ofpV = 0.06 − 0.08 (Bauer et al. 2013; Nugent et al. 2016). Nodetection was seen in STEREO HI-1A observations (limit-ing magnitude of m ∼ 13.5) near 1I/‘Oumuamua’s perihelionpassage at 0.25 au on 2017 September 9 (K. Battams, pers.comm). Consistent with this earlier non-detection, it was apoint source in deep VLT imaging on 2017 October 24, withno coma (Meech et al. 2017), and upper limits of surfacebrightness of 28–30 mag arcsec−2 within ∼ 5′′ radial distancewere set by Ye et al. (2017) on October 26 and Knight et al.(2017) on October 30. Additionally, no meteor activity fromassociated dust was seen (Ye et al. 2017). This implies ob-servations of 1I/‘Oumuamua directly measure its surface.

1 JPL Horizons heliocentric elements, as of 2017 November 14: https://ssd.jpl.nasa.gov/sbdb.cgi?sstr=2017%20U1

2 See footnote 1.

Measurement of the surface reflectivity of 1I/‘Oumuamuawill provide the first ever comparison between solar plan-etesimals and those from another star. Such measurementscould be used to infer the formation environment of this ob-ject, or provide evidence for a surface composition that isdistinct from solar planetesimals. However, 1I/‘Oumuamua’ssurface composition may have experienced substantial alter-ation during its exposure of Myr, and potentially Gyr, in in-terstellar space. No star has yet been confirmed as a poten-tial origin (Mamajek 2017), therefore the upper bound on1I/‘Oumuamua’s age is around 10 Gyr, after the formationof stars of moderate metallicity.

Compositional information on minor planets as small as1I/‘Oumuamua is limited. In reflectance relative to the colourof the Sun, larger Solar System objects range from neu-tral to substantially more red (e.g Jewitt 2015, and refer-ences therein), with spectra ranging from featureless (C-type asteroids) to strong absorption bands (S-type asteroids)(Rivkin et al. 2015; Reddy et al. 2015). Initial observationsof 1I/‘Oumuamua with 4–5-m class telescopes show a fea-tureless spectral slope that is similar to many small trans-Neptunian objects (TNOs): moderately redder than solar.Optical spectra in the wavelength range 400-950 nm from2017 October 25 and 26 include slopes of 30±15% (Masiero2017), 17±2% (Fitzsimmons et al. 2017), and 10±6% (Yeet al. 2017) per 1000 angstroms.

Both optical and near-infrared spectral information be-yond ∼ 1 µm are necessary to distinguish the compositionalclasses seen in the outer Solar System (Fraser & Brown2012; Dalle Ore et al. 2015; Pike et al. 2017). Thus, J-band photometry is key for establishing the relationshipof 1I/‘Oumuamua’s surface type to the Solar System. Wepresent near-simultaneous grJ photometry and colors of1I/‘Oumuamua in the optical and near-infrared, and com-pare to the colors of known Solar System bodies.

2. OBSERVATIONS AND ANALYSIS

2.1. Observations

We observed 1I/‘Oumuamua with two telescopes on 2017October 29. First, we observed 1I/‘Oumuamua with the 8.1-m Frederick C. Gillett Gemini North Telescope on Mau-nakea, during 05:50–07:55 UT. JPL Horizons3 predictedfrom the available 15-day arc that 1I/‘Oumuamua was thenat a heliocentric distance of 1.46 au and geocentric distanceof 0.53 au (phase angle α = 24.0◦), producing a rapid rate ofon-sky motion of R.A. 160 ′′/hr and Decl. 14 ′′/hr. We there-fore tracked the telescope non-sidereally at 1I/‘Oumuamua’srates (Fig. 1). The observations were in photometric skiesbetween airmass 1.04-1.14, with seeing in r′ of 0.7′′ to 0.5′′.

3 https://ssd.jpl.nasa.gov/horizons.cgi

COL-OSSOS: COLORS OF 1I/‘OUMUAMUA 3

The waxing 63%-illuminated Moon was 39◦ away, produc-ing a sky brightness of ∼ 20 mag/arcsec2 in r′.

g′, r′, and J imaging were obtained using the imagingmode of the Gemini Multi-Object Spectrograph (GMOS-N, Hook et al. 2004) and the Near-Infrared Imager (NIRI;Hodapp et al. 2003). The predicted apparent magnitude of1I/‘Oumuamua was mv = 22.7; exposure times are in Table 1.We observed with GMOS in two filters: r_G0303 (λ=6300 Å,δλ=1360 Å) and g_G0301(λ=4750 Å, δλ=1540 Å), whichwe refer to as r′ and g′, and which are similar to r and gin the Sloan Digital Sky Survey (SDSS) photometric system(Fukugita et al. 1996). GMOS was configured with the up-graded red-sensitive CCDs from Hamamatsu Photonics. Thetarget was kept on the middle GMOS CCD, and 2×2 binningwas used, resulting in an effective pixel scale of 0′′.1614.NIRI J band (λ=12500 Å, 11500-13300 Å coverage) imageswere acquired using the f/6 camera (pixel scale of 0′′.116).For both instruments, we dithered between exposures in thesame filter.

Any significant magnitude changes due to rotational vari-ability of 1I/‘Oumuamua will affect its measured colors.Our observing program therefore employed the design ofthe Colours of the Outer Solar System Origins Survey (Col-OSSOS; full detail of the techniques used will appear inSchwamb et al. 2017). Col-OSSOS uses repeated measure-ments in each band to distinguish any light curve effects fromthe change in surface reflectance during the observing se-quence, bracketing the NIRI observations with GMOS imag-ing in a r′g′Jg′r′ filter pattern. With this cadence, we canidentify if 1I/‘Oumuamua is variable on the timescale of ourobservations, and apply a correction, assuming that all filtersare similarly affected (see § 2.3). This is reasonable as bright-ness variations for solar system objects in 1I/‘Oumuamua’ssize regime are due to object shape; use of filter bracket-ing during Col-OSSOS has been effective at removing lightcurve effects (Pike et al. 2017).

We acquired specific observations for calibration tempo-rally adjacent to our science data. The optical sequence wasbracketed by a single 150 s sidereally-tracked exposure ineach optical filter, centered close to 1I/‘Oumuamua’s pre-dicted location. For the NIR calibration, we observed NIRstandard GD 246 at two different airmasses, with a set ofnine dithered, sidereally tracked exposures at the beginningand end of our observing sequence.

Thirteen hours after the Gemini observations, we observed1I/‘Oumuamua, then at very similar geometry with a phaseangle of α = 24.4◦, with the 4.2 m William Herschel Tele-scope (WHT) on La Palma from 19:45 UT – 21:52 UT. Non-sidereally guided imaging was obtained in photometric con-ditions and ∼ 1′′ seeing, at airmass 1.3–1.1, with the imagingmode of the ACAM imager/spectrograph (Benn et al. 2008)(pixel scale of 0′′.253). The data were acquired with four fil-

0.26 0.27 0.28 0.29 0.30 0.31 0.32 0.33MJD + 58055

22.0

22.1

22.2

22.3

22.4

22.5

22.6

22.7

22.8

App

aren

tMag

nitu

de

r’g’J

Color-corrected Gemini photometry of 1I/‘Oumuamua in g’, r’, and J(Table 1). The line indicates best fit.

GMOS-N r-band NIRI J-band

5’’ 5’’

Imaging of 1I/‘Oumuamua with non-sidereal tracking on the R.A. 160′′/hrmotion rates of 1I/‘Oumuamua; a single 300 s GMOS exposure in r (left)and a stack of 11 NIRI 120-s exposures in J (right). 1I/‘Oumuamua was

free of cosmic rays in all exposures.

Figure 1. Observations and photometry of 1I/‘Oumuamua fromGemini North on 2017 October 29.

ters: ING Filter #701, #702, #703 and #704, correspondingto g,r, i,z in the SDSS photometric system. Individual expo-sures were 100 s, in a sequence 6r−6g−7r−10i−6r−12z−6r(Table 1), with the repetition of r to test for variability.

2.2. Data Reduction

The Gemini and WHT images4 were prepared for analy-sis using standard reduction techniques. The GMOS obser-vations were bias-subtracted and flat-fielded using standardmethods with Gemini IRAF v1.14 in AstroConda, with amaster flat frame built from the last month of GMOS twi-

4 All reduced data are available at http://apps.canfar.net/storage/list/ColOSSOS/Interstellar.

4 BANNISTER ET AL.

light flats, and the CCD chips mosaiced into a single exten-sion. For the NIRI sequence, flat-fielding was with a masterflat frame built from the Gemini facility calibration unit flats.A sky frame was produced for each image of 1I/‘Oumuamuafrom the unshifted 14 frames closest in time to the image, andsubtracted from that image. We removed cosmic rays fromeach NIRI image with L. A. Cosmic (van Dokkum 2001),then aligned and stacked 20 of the 21 science 120-s NIRIexposures, rejecting one frame where 1I/‘Oumuamua was infront of a background source. To assess for any variability in1I/‘Oumuamua, we built independent stacks from each thirdof the J image sequence (Table 1). The WHT imaging werebias-subtracted and flatfielded with archival sky flats from2017 August, as the sky flats taken prior to the observationshad strong gradients across the field due to scattered moon-light and twilight variations.

Photometric measurements on all Gemini data (Table 1)were performed with TRIPPy (Fraser et al. 2016), using around aperture with a radius of 3.0× the point-spread func-tion (PSF)’s full-width-half-maximum (FWHM). Because ofthe non-sidereal tracking of our target, the elongated starsin the 1I/‘Oumuamua images cannot be used to computea point-source aperture correction. We instead derived amean stellar PSF profile from the calibrator images for eachdataset, and used this profile to calculate both the FWHM ra-dius and an aperture correction, for photometry on the non-sidereal 1I/‘Oumuamua images. For the optical GMOS pho-tometry, the four bracketing sidereally-tracked images werecalibrated to the SDSS system. Image zeropoints and a linearcolor term were fit using the instrumental SDSS magnitudesof stars in the calibration images, to scale the SDSS systemto the Gemini filter system. The zeropoints of the science im-ages were set from those of the calibrator temporally closestto each science image, introducing 0.02 magnitude of uncer-tainty in the calibration of those frames to encompass the sizeof the variation in the zeropoint. An additional 0.02 magni-tude uncertainty is due to the indirect measure of the aperturecorrection (Fraser et al. 2016). For the J-band NIRI photom-etry, we apply a mean aperture correction of 0.03 mag on the1I/‘Oumuamua data. While this does not account for seeingvariation along the longer J-band sequence, the highly stablesky conditions during our observations and the use of a largeaperture mitigate the need for a variable aperture correction.The magnitudes of 1I/‘Oumuamua in Table 1 from the Gem-ini observations are thus in the Gemini filter system.

Photometry on the WHT imaging (Table 1) used a slightlydifferent analysis, as we did not have sidereally tracked cal-ibrator frames. The photometry of trailed stars in all im-ages were measured using TRIPPy pill-shaped apertures,with length equal to the known rate of motion during theimage. Stellar centroids were found with Source Extrac-tion and Photometry in Python (Barbary et al. 2017), using

a custom linear kernel of the trail length and angle, whichwas convolved with a gaussian to simulate the appropriatestellar shape. A filter-dependent FWHM of 0.7–1.3′′ weremeasured directly from 1I/‘Oumuamua. It was not visible inthe z stack or in the fourth r stack, and thus we do not re-port z photometry. Apertures 7 pixels in radius (1.8′′) wereused for both the round aperture used to measure the fluxfrom 1I/‘Oumuamua, and the pill apertures used on the stars.As this implicitly assumes identical aperture corrections forboth aperture shapes, we adopt 0.02 magnitudes as a conser-vative estimate of measurement uncertainty, induced by theuse of a fixed aperture. Stellar calibration magnitudes wereextracted from the Pan-STARRS1 catalog (Chambers et al.2016), and converted to SDSS using the Tonry et al. (2012)transformations. As insufficient stars were available to mea-sure a color term, we required the SDSS stars to have similarcolors to 1I/‘Oumuamua: 0.1 < (g − r) < 0.7. This induceda 0.02 magnitude uncertainty in calibration. The magnitudesof 1I/‘Oumuamua in Table 1 from the WHT observations arethus in the SDSS system.

2.3. Color Computation

For the g − r and r − J colors of 1I/‘Oumuamua, a best-fitline and an average g − r color term were fit to the higher-precision Gemini optical data, in the Gemini system (Fig. 1).A mean r − J color was found by estimating the r value fromthe fitted line at each J stack epoch. Uncertainties in the fit,and in the individual J measurements were folded together.The g−r and r−J colors thus determined were then convertedto the SDSS system using the color terms determined fromthe initial calibration, carrying uncertainties in the color termappropriately.

The colors from the WHT measurements are consistent,but substantially more uncertain, and we subsequently con-sider only those from Gemini. We note the r − i color cor-responds to a spectral slope of 22±15%, consistent with theearlier reports. The measured SDSS colors of 1I/‘Oumuamuaare in Table 2.

3. VARIABILITY AND SHAPE OF 1I/‘OUMUAMUA

There are significant and correlated brightness increasesin optical and NIR for 1I/‘Oumuamua during our Geminiobservations, with the variability in J-band tracking that inthe r′ and g′ bands (Fig. 1). Over the 0.48 hours of J-band imaging, 1I/‘Oumuamua brightens systematically by0.183±0.065 magnitudes, and likewise, brightens by 0.66±0.03 and 0.71±0.05 magnitudes between the first and last r’and g’ observations, which respectively span 1.54 and 1.19hours. These are significant variations given the short time-frame (Table 1). As they correlate across filters, the bright-ening is most likely due to 1I/‘Oumuamua’s shape ratherthan to variability in albedo or surface spectral reflectance.

COL-OSSOS: COLORS OF 1I/‘OUMUAMUA 5

Table 1. Photometry of 1I/‘Oumuamua with the 8.1-m Frederick C.Gillett Gemini North Telescope and the 4.2 m William Herschel Tele-scope

MJD Filter Effective m f ilter NoteExposure (s)

GMOS-N and NIRI (Gemini North)

58055.25860 r_G0303 300 22.74±0.03

58055.26323 r_G0303 300 22.66±0.03

58055.26902 g_G0301 300 23.11±0.07

58055.27337 g_G0301 300 23.02±0.06

58055.28856 J 840 21.19±0.07 7-image stack

58055.29954 J 840 21.29±0.08 7-image stack

58055.30914 J 720 20.99±0.07 6-image stack

58055.31881 g_G0301 300 22.40±0.03

58055.32281 r_G0303 300 22.08±0.02

ACAM (WHT)

58055.82845 r_ING702 600 22.47±0.09 6-image stack

58055.83740 g_ING701 600 23.28±0.30 6-image stack

58055.84646 r_ING702 700 22.83±0.11 7-image stack

58055.86336 i_ING703 1000 22.81±0.08 10-image stack

58055.87835 r_ING702 600 23.40±0.18 6-image stack

NOTE—GMOS-N (Gemini) photometry is color term corrected, in the Gemini-Hamamatsu system. ACAM (WHT) photometry is in the SDSS system. Forboth, shot noise, calibration, and aperture correction uncertainties are incorpo-rated in quadrature.

Table 2. Optical-NIR SDSS Colors of1I/‘Oumuamua

Filters Measured Color Observations

g − r 0.47±0.04 Gemini

g − r 0.63±0.31 WHT

r − i 0.36±0.16 WHT

r − J 1.20±0.11 Gemini

NOTE—Gemini g− r and r −J colors are near-simultaneous; WHT g − r and r − i are from13 hours later.

This is supported by the general consistency of the WHTcolours, observed at a different part of 1I/‘Oumuamua’slightcurve (Fig. 2, discussed below). Our image stacks allhad point-like PSFs, and we thus do not assess upper limitsfor the presence of coma. However, the overall periodicityof 1I/‘Oumuamua’s brightness (Fig. 2) also implies our ob-served increase in brightness in the Gemini data is not fromdust emission.

We assess the lightcurve of 1I/‘Oumuamua by combin-ing our optical photometry from 2017 October 29 (Table 1)with the optical photometry of Knight et al. (2017) with the4.3 m Discovery Channel Telescope (DCT) on 2017 Octo-ber 30. The combined, geometrically corrected photome-try is in Fig. 2. We analyze the photometry for periodicityusing both the Lomb-Scargle technique (Lomb 1976) and amodified phase-dispersion minimisation (PDM) fitting tech-nique (Stellingwerf 1978). Instead of the typical PDM modelwhich bins the data and looks for the place where the pointsin the bins are not as dispersed as other periods, our modi-fied PDM goes through every possible period and folds thedata, then fits a second-order Fourier series to each foldedlightcurve. The quality of fit is calculated from the residuals,after which the best fit is chosen (M. W. Buie, pers. comm.).We obtain a consistent double-peaked period of 8.10± 0.42hours, with a peak-to-peak amplitude from the fitted modelof ∆m = 1.8 magnitudes. We note that the last pair of Geminiobservations imply an excursion from our lightcurve model;the data are entirely reliable, given the consistent stabilityof the observing conditions and the calibrations, and fromcareful inspection of the images. Our results using only ourGemini and WHT photometry with that from the DCT areindependently in agreement with those of Bolin et al. (2017),which used photometry from the 3.5 m Apache Point Obser-vatory and the DCT photometry.

Its lightcurve implies that 1I/‘Oumuamua is either a veryelongated object, or a contact binary system of two equalsized, prolate components aligned for maximum elongation(Sheppard & Jewitt 2004; Leone et al. 1984). From simula-tions for resolved Centaur binaries in our Solar System (Nollet al. 2006), a contact binary would stay intact through per-ihelion. The light curve yields consistent results in eithercase. We consider 1I/‘Oumuamua’s elongation and densityassuming it is a prolate ellipsoid with semi-axes a > b = c.The observed ∆m = 1.8 mags would require an axis ratioof a/b = 5.3, or larger if 1I/‘Oumuamua was not observedequator-on (Lacerda & Luu 2003). Such an ellipsoid spin-ning in P = 8.1 hours with a/b = 5.3 (Fig. 2) would require adensity at least ρ = (a/b)2(3π)/(GP2) = 5.9 g.cm−3 to preventit from shedding regolith, consistent with the observed ab-sence of coma. If it is instead a contact binary of two prolatecomponents, each with axes ratio 0.5(a/b) (to produce thesame ∆m) a similar density of 5.9 g.cm−3 is required to hold

6 BANNISTER ET AL.

0.0 0.2 0.4 0.6 0.8 1.0

22

23

24

22

23

24

Rotational Phase

AbsoluteMagnitude

Gemini r Gemini g WHT DCT

Figure 2. Lightcurve of 1I/‘Oumuamua during 2017 October 29–30. Absolute magnitude assumes a linear phase function with slope 0.03mag/deg. Rotational phase starts at MJD 58055.0 and assumes a spin period P = 8.1 hr. Overplotted is a model based on a prolate ellipsoid withaxes ratio a/b = 5.3, which has ∆m = 1.8 mag. A similar curve would be produced by a contact binary with equal sized, prolate componentseach with a/b ∼ 2.7, elongated along a line connecting their centers.

the components in mutual orbit. As these densities are unrea-sonably higher than those of likely compositions of silicateor icy materials, it requires that 1I/‘Oumuamua has internalstrength.

Small TNOs with H < 9 and Centaurs with H < 11 typi-cally have 7–9 hour rotation periods with peak-to-peak vari-ations of ∼0.3 magnitudes (Duffard et al. 2009), though lightcurve amplitude changes of a magnitude or more have beenmeasured for TNOs in this size range (Benecchi & Shep-pard 2013). Small asteroids with 1I/‘Oumuamua’s degreeof elongation are rare but not unknown; examples includethe ∼ 200−300 m diameter Near-Earth Asteroids 2001 FE90and 2007 MK13, both with lightcurve amplitudes ≥ 2.1 mag-nitudes (Warner et al. 2009).

4. 1I/‘OUMUAMUA IN CONTEXT WITH THE COLORSOF THE MINOR PLANETS OF THE SOLAR SYSTEM

For comparison with 1I/‘Oumuamua, we collated colorsof minor planet populations in our Solar System from fourdatasets of optical and near-IR measurements: SMASS forasteroids, Emery et al. (2011); Marsset et al. (2014) forJupiter Trojans, MBOSS for a variety of distant populations,and Col-OSSOS for trans-Neptunian objects (TNOs). The

Small Main-Belt Asteroid Spectroscopic Survey (SMASS)5

(Xu et al. 1995; Burbine & Binzel 2002; Bus & Binzel2002b) measured spectra for a range of asteroid dynami-cal groups, from near-Earth objects through the main belt toMars-crossing asteroids. SMASS provided 157 objects withboth optical and NIR spectra, representing 23 of the 26 spec-tral types in the Bus & Binzel (2002a) taxonomy (the Cg, D,and Q types were not available in NIR). For P and D types,we used the 38 Jupiter Trojans with NIR spectra from Emeryet al. (2011) for which Marsset et al. (2014) collated opticalspectra. We converted each spectra to grJ colors by convolv-ing the filter bandpasses. The mean and range for each spec-tral type are shown in Fig. 3. The Minor Bodies in the OuterSolar System (MBOSS) (Hainaut et al. 2012) database6 in-dexes the reported colors for objects in outer Solar Systemdynamical populations, including Jupiter Trojans, short- andlong-period comets, Centaurs and TNOs. 47 objects in theMBOSS tabulation have measurements in comparable filtersand of sufficiently high-SNR to provide comparison to ourgrJ measurements of 1I/‘Oumuamua. The MBOSS colorswere converted to g − r and r − J assuming a linear spectrumthrough the V,R,g,r range. We retained objects if the uncer-

5 http://smass.mit.edu/smass.html6 http://www.eso.org/~ohainaut/MBOSS/

COL-OSSOS: COLORS OF 1I/‘OUMUAMUA 7

0.4 0.6 0.8 1.0 1.2 1.4g'-r'

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2r'-

J

S-type asteroidsC-type asteroidsX-type asteroidsother asteroidsJupiter TrojansTNOs/JFCs (MBOSS)TNOs (Col-OSSOS)1I/2017 U1Sun

Figure 3. The grJ colors of 1I/‘Oumuamua in context with the known Solar System. The mean color and range of asteroids and the twocolor classes of Jupiter Trojans are given from 157 asteroids in 23 of the 26 Bus & Binzel (2002a) spectral types as recorded in SMASS, and38 Jupiter Trojans (Emery et al. 2011; Marsset et al. 2014). The much more sparsely sampled distant populations are shown with individualobjects, with their measurement uncertainties from MBOSS and Col-OSSOS. The observations of 1I/‘Oumuamua and of Col-OSSOS wereperformed in the same way.

8 BANNISTER ET AL.

tainties on their color measurements had d(V − R) < 0.4 andd(V − J) < 0.4. Col-OSSOS provides high-precision grJ col-ors for mr < 23.6 TNOs from the Outer Solar System OriginsSurvey (Bannister et al. 2016), acquired in the same manneras our observations of 1I/‘Oumuamua with Gemini (§ 2.1).We use the 21 Col-OSSOS TNOs with grz photometry dis-cussed in Pike et al. (2017); the grJ measurements are forth-coming in Schwamb et al. (2017).

The grJ colors of 1I/‘Oumuamua are at the neutral endof the Solar System populations (Fig. 3). About 15% ofthe trans-Neptunian objects have colors consistent with1I/‘Oumuamua, all in dynamically excited populations.1I/‘Oumuamua’s color is also consistent with that of the lessred Jupiter Trojans, which are P type (Emery et al. 2011),and with Bus & Binzel (2002a); DeMeo et al. (2009) X typein the asteroids, which encompasses the Tholen (1984) E, Mand P classifications. As its albedo is unknown, we do notdescribe 1I/‘Oumuamua as consistent with Tholen (1984) Ptype.

Notably, 1I/‘Oumuamua does not share the distinctly red-der colors of the cold classical TNOs (Tegler et al. 2003; Pikeet al. 2017), which may be on primordial orbits. Nor is itscolor among the red or “ultra-red" colors of the larger TNOson orbits that cross or are well exterior to the heliopause(Sheppard 2010; Trujillo & Sheppard 2014; Bannister et al.2017). The cause of ultra-red coloration of these TNOs isunknown, but has been attributed to long-term cosmic ray al-teration of organic-rich surfaces (Jewitt 2002), such as wouldbe expected during the long duration of interstellar travel.

While this work was under review, several other well-constrained color and spectral measurements were reported.Our optical color is compatible with that observed in BV R byJewitt et al. (2017), but 3σ discrepant from the mean g − r =0.84±0.05 over multiple nights of Meech et al. (2017). Ourmeasurements confirm consistent color over roughly a quar-ter of 1I/‘Oumuamua’s surface, so the discrepancy likely in-dicates a change in surface color elsewhere on the body. Thisis supported by considering spectral slope through the near-IR; our r − J corresponding to 3.6%/100nm is more neu-tral than the 7.7 ± 1.3%/100nm observed by Fitzsimmonset al. (2017) over 0.63µm to 1.25µm. Note that our 1.15–1.33µm J-band data is at longer wavelengths than the Meechet al. (2017) Y -band (0.97–1.07µm), so we make no di-rect comparison there. More extensive modelling of surfacecolor patchiness and non-principal axis rotation (tumbling)of 1I/‘Oumuamua is considered by Fraser et al. (2017).

1I/‘Oumuamua’s largely neutral color opens up a numberof possibilities. It could imply that the correlation of ultra-redness with heliocentric distance has an alternative cause. Itcould suggest that 1I/‘Oumuamua formed with an organics-poor surface, within its star’s water ice line. 1I/‘Oumuamua’scolor being within the observed range for minor planets in

the Solar System could support that 1I/‘Oumuamua origi-nated from a star from the Sun’s birth cluster, which shouldhave a similar chemistry. A possible additional complicationcould be resurfacing due to surface activity, which wouldaffect surface color. This seems unlikely as no surface ac-tivity was detected during 1I/‘Oumuamua’s perihelion pas-sage, but 1I/‘Oumuamua could have had past activity in itsorigin system or in another close encounter. We empha-size that our observations only probe the top few micronsof 1I/‘Oumuamua’s surface.

Our Gemini and WHT observations provide high-precision,lightcurve-independent optical and NIR color measurementsfor 1I/‘Oumuamua. With its period of 8.1±0.4 hours, highlyelongated ≥ 5.3 : 1 ellipsoidal or prolate-binary shape, andneutral grJ color, 1I/‘Oumuamua is within the known pa-rameters of minor planets from the Solar System, but lies atthe extreme ends of the physical ranges.

The authors acknowledge the sacred nature of Maunakeaand appreciate the opportunity to observe from the mountain.This work is based on observations from Director’s Discre-tionary Time (DDT) Program GN-2017B-DD-8, obtained atthe Gemini Observatory, which is operated by the Associa-tion of Universities for Research in Astronomy, Inc., under acooperative agreement with the NSF on behalf of the Gem-ini partnership: the National Science Foundation (UnitedStates), the National Research Council (Canada), CONICYT(Chile), Ministerio de Ciencia, Tecnología e Innovación Pro-ductiva (Argentina), and Ministério da Ciência, Tecnologiae Inovação (Brazil). The authors thank the queue coordina-tors, NIRI and GMOS instrument teams, science operationsspecialists, and other observatory staff at Gemini North fortheir support of our DDT program and their support of theCol-OSSOS program. We also thank Andy Stephens for hisassistance during the Gemini observations.

The WHT is operated on the island of La Palma by theIsaac Newton Group of Telescopes in the Spanish Observa-torio del Roque de los Muchachos of the Instituto de As-trofísica de Canarias. The ACAM data were obtained as partof programme SW2017b11. We thank Ian Skillen for advis-ing on and performing the WHT observations.

The authors greatly thank Atsuko Nitta for her supportof the Gemini observing program and for being a soundingboard for program ideas and observing strategies. We espe-cially acknowledge and thank the online planetary commu-nity on Twitter for productive discourses and sharing of pre-liminary results related to 1I/‘Oumuamua that helped spurthese observations.

M.T.B. appreciates support from UK STFC grant ST/L000709/1.M.E.S., A.B.S. and S.S. were supported by Gemini Observa-tory.

COL-OSSOS: COLORS OF 1I/‘OUMUAMUA 9

This research has made use of NASA’s Astrophysics DataSystem Bibliographic Services, the JPL HORIZONS web in-terface (https://ssd.jpl.nasa.gov/horizons.cgi), and data and services provided by the InternationalAstronomical Union’s Minor Planet Center.

Facilities: Gemini:Gillett (GMOSN, NIRI), ING:Herschel(ACAM)

Software: astropy (The Astropy Collaboration et al.2013), TRIPPy (Fraser et al. 2016), SExtractor (Bertin& Arnouts 1996), SEP (Barbary et al. 2017), Astro-Conda (http://astroconda.readthedocs.io/,L.A. Cosmic (van Dokkum 2001))

REFERENCES

Bacci, P., Maestripieri, M., Tesi, L., et al. 2017, M.P.E.C.2017-U181. Comet C/2017 U1 (PANSTARRS), , .http://www.minorplanetcenter.net/mpec/K17/

K17UI1.html

Bannister, M. T., Kavelaars, J. J., Petit, J.-M., et al. 2016, AJ, 152,70

Bannister, M. T., Shankman, C., Volk, K., et al. 2017, AJ, 153, 262Barbary, K., Boone, K., Craig, M., Deil, C., & Rose, B. 2017,

doi:10.5281/zenodo.896928Barclay, T., Quintana, E. V., Raymond, S. N., & Penny, M. T.

2017, ApJ, 841, 86Bauer, J. M., Grav, T., Blauvelt, E., et al. 2013, ApJ, 773, 22Benecchi, S. D., & Sheppard, S. S. 2013, AJ, 145, 124Benn, C., Dee, K., & Agócs, T. 2008, in Proc. SPIE, Vol. 7014,

Ground-based and Airborne Instrumentation for Astronomy II,70146X

Bertin, E., & Arnouts, S. 1996, Astronomy and AstrophysicsSupplement, 117, 393

Bolin, B., Weaver, H., Fernandez, Y., & Lisse, C. 2017, Submittedto AJ, arXiv:1711.04927

Burbine, T. H., & Binzel, R. P. 2002, Icarus, 159, 468Bus, S. J., & Binzel, R. P. 2002a, Icarus, 158, 146—. 2002b, Icarus, 158, 106Chambers, K. C., Magnier, E. A., Metcalfe, N., et al. 2016, AJ

submitted, arXiv:1612.05560Cook, N. V., Ragozzine, D., Granvik, M., & Stephens, D. C. 2016,

ApJ, 825, 51Dalle Ore, C. M., Barucci, M. A., Emery, J. P., et al. 2015, Icarus,

252, 311DeMeo, F. E., Binzel, R. P., Slivan, S. M., & Bus, S. J. 2009,

Icarus, 202, 160Duffard, R., Ortiz, J. L., Thirouin, A., Santos-Sanz, P., & Morales,

N. 2009, Astronomy and Astrophysics, 505, 1283Emery, J. P., Burr, D. M., & Cruikshank, D. P. 2011, AJ, 141, 25Engelhardt, T., Jedicke, R., Vereš, P., et al. 2017, AJ, 153, 133Fitzsimmons, A., Snodgrass, C., Rozitis, B., et al. 2017, in pressFrancis, P. J. 2005, ApJ, 635, 1348Fraser, W., Alexandersen, M., Schwamb, M. E., et al. 2016, AJ,

151, 158

Fraser, W. C., & Brown, M. E. 2012, ApJ, 749, 33Fraser, W. C., Pravec, P., Fitzsimmons, A., et al. 2017, submitted,

arXiv:1711.11530Fukugita, M., Ichikawa, T., Gunn, J. E., et al. 1996, AJ, 111, 1748Hainaut, O. R., Boehnhardt, H., & Protopapa, S. 2012, A&A, 546,

A115Hodapp, K. W., Jensen, J. B., Irwin, E. M., et al. 2003, PASP, 115,

1388Hook, I. M., Jørgensen, I., Allington-Smith, J. R., et al. 2004,

PASP, 116, 425Jewitt, D. 2015, AJ, 150, 201Jewitt, D., Luu, J., Rajagopal, J., et al. 2017, ApJL, 850, L36Jewitt, D. C. 2002, AJ, 123, 1039Knight, M. M., Protopapa, S., Kelley, M. S. P., et al. 2017, ApJL,

arXiv:1711.01402Lacerda, P., & Luu, J. 2003, Icarus, 161, 174Leone, G., Paolicchi, P., Farinella, P., & Zappala, V. 1984, A&A,

140, 265Levison, H. F., Duncan, M. J., Brasser, R., & Kaufmann, D. E.

2010, Science, 329, 187Lomb, N. R. 1976, Ap&SS, 39, 447Mamajek, E. 2017, Research Notes of the American Astronomical

Society, 1, 21Marsset, M., Vernazza, P., Gourgeot, F., et al. 2014, A&A, 568, L7Masiero, J. 2017, submitted to ApJL, arXiv:1710.09977Meech, K., Bacci, P., Maestripieri, M., et al. 2017, M.P.E.C.

2017-U183. A/2017 U1, , . http://www.minorplanetcenter.net/mpec/K17/K17UI3.html

Meech, K. J., Weryk, R., Micheli, M., et al. 2017, Nature, 1Noll, K. S., Levison, H. F., Grundy, W. M., & Stephens, D. C.

2006, Icarus, 184, 611Nugent, C. R., Mainzer, A., Bauer, J., et al. 2016, AJ, 152, 63Pike, R. E., Fraser, W. C., Schwamb, M. E., et al. 2017, AJ, 154Reddy, V., Dunn, T. L., Thomas, C. A., Moskovitz, N. A., &

Burbine, T. H. 2015, in Asteroids IV, ed. P. Michel, F. E.DeMeo, & W. F. Bottke (University of Arizona Press), 43–63

Rivkin, A. S., Campins, H., Emery, J. P., et al. 2015, in AsteroidsIV, ed. P. Michel, F. E. DeMeo, & W. F. Bottke (University ofArizona Press), 65–87

10 BANNISTER ET AL.

Schwamb, M. E., Fraser, W., Bannister, M. T., et al. 2017, inpreparation

Sheppard, S. S. 2010, AJ, 139, 1394Sheppard, S. S., & Jewitt, D. 2004, AJ, 127, 3023Stellingwerf, R. F. 1978, ApJ, 224, 953Tegler, S. C., Romanishin, W., & Consolmagno, G. J. 2003, ApJL,

599, L49The Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al.

2013, Astronomy and Astrophysics, 558, A33Tholen, D. J. 1984, PhD thesis, University of Arizona, Tucson

Tonry, J. L., Stubbs, C. W., Lykke, K. R., et al. 2012, ApJ, 750, 99

Trujillo, C. A., & Sheppard, S. S. 2014, Nature, 507, 471

van Dokkum, P. G. 2001, PASP, 113, 1420

Warner, B. D., Harris, A. W., & Pravec, P. 2009, Icarus, 202, 134

Xu, S., Binzel, R. P., Burbine, T. H., & Bus, S. J. 1995, Icarus, 115,

1

Ye, Q.-Z., Zhang, Q., Kelley, M. S. P., & Brown, P. G. 2017, ApJL,

arXiv:1711.02320